Neurotechnology releases brainwave biometrics development kit and updated image recognition platform

Neurotechnology has updated its deep learning image labeling and object recognition platform SentiSight.ai, and also launched a development kit for electroencephalography (EEG) biometrics, according to a pair of company announcements.

SentiSight.ai now features model training for object detection, offline models, project sharing and improvements to the image labeling tool. The web-based platform is intended to make image annotation as convenient and efficient as possible, even with many people working together on large projects, and to provide a user-friendly user interface for both training and deploying deep neural network models, the company says. The features enable models to be trained and improved in an iterative process.

The new version also provides similarity searches and time tracking for labeling projects.

“SentiSight.ai has become a platform of choice for image labeling and almost any AI-related task,” comments Dr. Karolis Uziela, SentiSight.ai team lead from Neurotechnology. “It has also become one of the first such platforms to offer the ability to download offline models, which allows our clients to be completely independent both from the platform and from their connection to the internet.”

The new BrainAccess Development Kit provides a full dry-contact EEG solution, including electrodes, headwear, electroencephalograph, software for EEG signal acquisition and processing, and brain-computer interface (BCI) example algorithms, according to the announcement.

Conventional EEG technology uses “wet” electrodes, requiring gel and scalp preparation for sufficient electrical contact, which makes it very challenging to use outside of a clinical or laboratory setting, Neurotechnology observes. The contact electrodes provided in the BrainAccess Development Kit are shape-conforming for superior comfort and electrical contact.

“EEG is a highly effective, low-cost method for measuring brain activity in BCI applications due to its non-invasive nature and high temporal resolution,” said Dr.Osvaldas Putkis, project lead at Neurotechnology. “However, two things are needed in order for EEG to become practical enough for widespread adoption and everyday BCI use: 1) efficient dry-contact electrodes for measurement and 2) miniaturization of electroencephalographs with a high number of acquisition channels.”

The company is also offering the BrainAccess MINI portable electroencephalograph with 16 acquisition channels, wireless connectivity and more than ten hours of battery life. The BrainAccess MODULAR electroencephalograph provides 32 to 128 acquisition channels in a modular architecture.

Standard and Extended versions of the Development Kit are offered, the former with the BrainAccess MINI or BrainAccess CAP with 16 electrodes. The latter includes BrainAccess MODULAR or CAP with 32 electrodes and expansion modules for up to 128.

BrainAccess software includes a motion classifier algorithm, an SSVEP (steady-state-evoked-potential) detector algorithm, and an alpha brainwave detector algorithm.

Article Topics

artificial intelligence  |  biometrics  |  brainwave  |  deep learning  |  image recognition  |  Neurotechnology